166 Journal of Vector Ecology June 2008

Colonization of abandoned swimming pools by larval mosquitoes and their predators following Hurricane Katrina

Kevin A. Caillouët1, John C. Carlson2, Dawn Wesson1, and Frank Jordan3

1Department of Tropical Medicine, Tulane University, New Orleans, LA 70112, U.S.A. 2Department of Pediatrics, Tulane University School of Medicine, New Orleans, LA 70112, U.S.A. 3Department of Biological Sciences, Loyola University New Orleans, New Orleans, LA 70118, U.S.A.

Received 6 July 2007; Accepted 20 January 2008

ABSTRACT: Thousands of flooded swimming pools were abandoned in New Orleans following Hurricane Katrina and provided a natural experiment to examine colonization of a novel aquatic habitat by mosquito larvae and their aquatic predators. We conducted a randomized survey of flooded swimming pools in two neighborhoods in January 2006 and found that 64% contained mosquito larvae, 92% contained predatory invertebrates, and 47% contained fishes. We collected 12,379 immature mosquitoes representing five species, primarily Culiseta inornata, and secondarily, the arboviral vector quinquefasciatus. Dragonfly nymphs in the families Aeshnidae and Libellulidae were the most common predatory invertebrates collected among a total of 32 non-mosquito invertebrate species. Eleven species of fishes were collected, with Gambusia affinis accounting for 76% of the catch. Diversity of fishes in swimming pools was positively correlated with proximity to a levee breach and the fish assemblage found in swimming pools was similar to that found along shorelines of Lake Pontchartrain and drainage canals that flooded the study area. Mosquito larvae were rare or absent from pools containing fishes; however, path analysis indicated that the presence of top predators or abundant competitors may somewhat mitigate the effect ofGambusia affinison mosquito presence. Journal of Vector Ecology 33 (1): 166-172. 2008.

Keyword Index: Swimming pools, mosquitoes, predators, Hurricane Katrina, path analysis.

INTRODUCTION return home and resume maintenance of their swimming pools within a few days. However, what happens to larval Tropical cyclones (hurricanes and typhoons) have pro- mosquito populations when large numbers of residents are found negative effects on natural and managed landscapes, unable to return home and flooded swimming pools are but they also provide unique ecological opportunities for necessarily abandoned for months or years? some species. For example, mosquito larvae and other An opportunity to address this question arose in aerially dispersing with aquatic immature stages are August 2005 when approximately 80% of New Orleans very successful at colonizing newly-flooded habitats (Durso was flooded as a result of levee failures associated with and Burguin 1988). Flood-induced changes may increase Hurricane Katrina (Reichardt et al. 2005). Brackish the populations of hematophagous but are not floodwater from Lake Pontchartrain inundated most areas generally associated with increases in vector-borne diseases of the city for up to three weeks. Flooding resulted in the in developed regions (Nasci and Moore 1998). However, creation of novel aquatic habitats that could be colonized flooding often contributes to increased biting by nuisance by opportunistic mosquito species. Furthermore, prolonged arthropods and may inhibit recovery and restoration of inundation could facilitate dispersal of fishes and other storm-ravaged landscapes (Shultz et al. 2005). Changes in aquatic organisms that resided in Lake Pontchartrain, the type and amount of larval habitat likely contributed to drainage canals, and recreational lagoons and ponds. We observed changes in size and composition of adult mosqui- carried out a survey of New Orleans swimming pools about to populations in New Orleans, LA, following the flooding four months after Hurricane Katrina in order to address associated with Hurricane Katrina (Caillouët et al. unpub- four related questions: Were mosquitoes using abandoned lished data). swimming pools for larval development? Were fishes able Aquatic habitats in flooded urban areas include to move across the flooded urban landscape and colonize lowlands, depressions, and artificial containers. Another abandoned swimming pools? Did aerially-dispersing type of novel aquatic habitat created by natural disasters are predators colonize abandoned swimming pools? Finally, abandoned swimming pools that are not maintained during were fishes and predatory insects able to regulate larval the recovery and restoration of urban landscapes. Abandoned mosquito populations in abandoned swimming pools? swimming pools harbor vectors of human disease (Carlson et al. 2004, Townsend 2005). There is little evidence that pools abandoned following tropical cyclones contribute to human disease risk because residents are typically able to Vol. 33, no. 1 Journal of Vector Ecology 167

MATERIALS AND METHODS by interfering with the direct predatory effects of the spe- cialist predator. This model was applied to the swimming Study area pools in this study, using presence of mosquito larvae as We surveyed swimming pools in the Lakeview and the “prey,” Western mosquitofish Gambusia( affinis) abun- Filmore neighborhoods in North Central New Orleans dance as the “specialist predators,” Odonata abundance as (Figure 1) (North of I-610, South of Robert E. Lee Blvd., the “competitors,” and presence of large predacious bugs East of the 17th Street Canal, and West of the London Ave. (Heteroptera: Belostomatidae and ) as “top preda- Canal) due to their proximity to breached levees and the tors” (Figure 3B). Path analysis (AMOS 4.0) was used to relative uniformity of flooding in this part of the city. Using derive standardized coefficients to describe the relative in- high-resolution aerial imagery captured before Hurricane direct and direct effects in this food web. Path analysis uses Katrina (Google Earth 4.0), we identified 270 swimming standardized coefficients from multiple regression analysis pools in Lakeview and 166 swimming pools in Filmore. to examine the relative strengths of various trophic interac- We randomly selected and visited 50 swimming pools tions within food web models (Miura and Takahashi 1988, from each neighborhood between 18th and 26th January Manickchand-Heileman et al. 2004). Abundances of Odo- 2006. We could not verify the existence of eight pools (8%) nata and G. affinis were log-transformed to meet assump- either because of misclassification or because they were tions of normality. above-ground pools that had been removed or destroyed. Seven pools (7%) could not be accessed because of locked RESULTS gates and another 21 pools were excluded from the survey because they had been drained (12%), cleaned (8%), or Sixty-nine percent (64/92 existing pools) of the covered (1%). We therefore sampled 64 of the 100 pools swimming pools in our study area remained abandoned originally selected, or 32 pools each from the Lakeview and four months after Hurricane Katrina. Of these pools, 64% Filmore neighborhoods (Figure 1). The mean surface area (41 pools) contained mosquito larvae, 92% (59 pools) of the sampled pools was 50.7 m2 ± 19.1 SD. contained predatory invertebrates, and 47% (30 pools) contained fishes. We collected 12, 379 immature mosquitoes Sampling methods representing five species (Table 1). Eighty-six percent of all We used a 300-ml dipper to sample mosquito larvae mosquitoes collected were identified as Culiseta inornata. (O’Malley 1989). Ten dips were collected haphazardly within 2 m of the edge of the pool and the contents of these dips were filtered through a 200 µm sieve and preserved in 70% ethanol. We used long-handled dip nets (mesh size = 6 mm, mouth area = 1400 cm2) to sample fishes and aquatic invertebrates (Figure 2) large enough to prey upon mosquito larvae (Turner and Trexler 1997). We made 40 1.5-m sweeps in each swimming pool and included all microhabitats (i.e., bottom, mid-water, surface, and near debris). Each sweep Industrial Canal sampled about 210,000 cm3. Organisms were removed from 17 th St. Canal each sweep sample and preserved in 10% buffered formalin. A In the laboratory, mosquito larvae (Fox and Wesson 2005) B

and other invertebrates were identified, enumerated, and Orleans Outfall Canal London Ave. Canal categorized by feeding guild (Merritt and Cummings 1996).

Data analysis Using Global Positioning System (GPS) coordinates, distances were calculated from each of the 32 Lakeview swimming pools to the 17th Street Canal levee breach, which was a likely source of fishes for the Lakeview neighborhood. We used Pearson correlation analysis to test for a relationship between fish species richness and distance to the 17th Street Canal levee breach. A model food web was created to evaluate the trophic effects of predator combinations on the presence of mosqui- to larvae (Figure 3A). This model is based on the hypoth- Figure 1. Map of study area and flooding extent, New esis that specialist predators (i.e., fish) are most efficient at Orleans, LA. A. Lakeview neighborhood. B. Filmore reducing prey (i.e. immature mosquitoes). Therefore, the neighborhood. Large arrows indicate direction of storm presence of top predators (i.e., belostomatids and nepids) surge from Lake Pontchartrain. Small arrows indicate levee and competitors (i.e., odonates) will indirectly benefit prey breach (source of flooding). 168 Journal of Vector Ecology June 2008

Culex quinquefasciatus (8%), Cx. salinarius (4%), Cx. tarsalis DISCUSSION (3%), and Anopheles atropos (0.1%) were also collected. We collected 1,364 non-mosquito invertebrates The floodwaters that covered 80% of New Orleans in representing 16 families (Table 1). Dragonfly nymphs of September 2005 did not fully recede for over three weeks, Families Libellulidae and Aeshnidae were the most common and the extent of destruction has kept many homeowners large-bodied invertebrates collected, occurring in 47% and from returning to the city over two years later (Plyer and 56% of pools, respectively. Bonaguro 2007). In January 2006, immature mosquitoes We collected 855 fishes belonging to 11 species (Table inhabited 64% of the abandoned swimming pools sampled. 2). The majority (76%) were G. affinis. Other numerically Most of these mosquitoes were larvae of the seasonally dominant fishes included sailfin molliesPoecillia ( latipinnia, abundant winter mosquito Cs. inornata, which is a 9%), least killifish (Heterandria formosa, 6%), and sheepshead laboratory vector of West Nile virus (WNV) (Goddard et minnows (Cyprinodon varieagatus, 5%). Small numbers of al. 2002), Saint Louis encephalitis virus, and Western equine inland silversides (Menidia beryllina), rainwater killifish encephalomyelitis virus (Hammon and Reeves 1943a, (Lucania parva), diamond killifish (Adinia xenica), bayou 1943b). The New Orleans Mosquito and Termite Control killifish Fundulus( pulverous), golden topminnows (Fundulus Board estimates that about 5,000 abandoned swimming chrysotus), naked gobies (Gobiosoma bosc), and introduced pools need long-term mosquito abatement (S. Sackett, Rio Grande cichlids (Cichlasoma cyanoguttatum) were also personal communication). The substantial number of collected. Species richness of fishes in 32 non-maintained potential vector mosquitoes we collected per pool and the Lakeview swimming pools was negatively correlated (r=- significant number of flooded abandoned swimming pools 0.404, P=0.022) with distance to the 17th Street Canal levee underscore the public health import of this newly-abundant breach. habitat. Path analysis revealed that G. affinis had a very large The unexpectedly high density of larval mosquitoes direct negative effect on the presence of mosquito larvae, in swimming pools is especially surprising given that this but that this effect was mitigated somewhat by the presence study was conducted in January, a month in which both of top predators and competitors (Figure 3B). We found mosquito activity and arbovirus transmission are typically no direct effects of Belostomatidae/Nepidae presence on very low (Michaels et al. 2005). The high productivity of Odonata abundance. Additionally, we found no direct swimming pools in January suggests that, if left untreated, effect of Odonata abundance on mosquito presence. A these habitats will be important for both vector and nuisance non-significant lack-of-fit test indicates there is agreement mosquitoes during warmer months. Though we collected between field data and the path model shown in Figure 3B predominantly Cs. inornata from the pools, the second most (χ2= 0.152, 1 df, p=0.696). abundant mosquito in our study was Cx. quinquefasciatus, the primary vector of WNV in Louisiana (Godsey et al. 2005). It is likely that as 2006 progressed, other mosquitoes, including Cx. quinquefasciatus, became more prolific in the

Figure 2. A flooded abandoned swimming pool in the Lakeview neighborhood of New Orleans. Vol. 33, no. 1 Journal of Vector Ecology 169

Table 1. Summary data for Hexapods collected from abandoned swimming pools in New Orleans in January 2006. Family groups are presented in bold along with their corresponding total organisms collected (also in bold).

# Organisms Order Family Species collected Belostomatidae 38 Belostoma lutarium 6 Belostoma testacium 31 Lethocercus uhleri 1 Gerridae Gerris canaliculatus 4 Nepidae 30 australis 29 Ranatra buenoi 1 Notenectidae 11 Buenoa margaritacea 2 Buenoa scimitra 2 indica 2 Notonecta irrorata 5 Corixidae Tricorixa louisianae 2 Naucoridae Pelocoris carolinensis 1 Odonata Libellulidae Pantala hymanaea 289 Aeshnidae Anax junius 327 Coenagrionidae Ishnura ramburii 174 Coleoptera 214 81 nimbatus 16 Tropisternus quadristriatus 12 Tropisternus blatchleyi 99 Dytiscidae 238 Agabus sp. 3 Cybister fimbriolatus crotchi 5 Graphoderus sp. 1 Rhantus sp. 1 Thermonectes basillaris 223 Thermonectes ornaticollis 5 Haliplidae Peltodytes sp. 12 Ephemeroptera Baetidae Callibaetis sp. 17 Odontomyia Stratiomyidae 9 Catatasima sp. 6 Diptera Odontomyia sp. 3 Syrphidae Eristalis sp. 2 Ephyridae Brachydeutera sp. 1 Noctuidae Archanura sp. 1 Culicidae 12,379 Culiseta inornata 4th instar 1,570 Culex quinquefasciatus 4th instar 138 Culex salinarius 4th instar 64 Culex tarsalis 4th instar 50 Anopheles atropos 4th instar 2 Unknown Pupae 655 3rd instar 3,084 2nd instar 3,799 1st instar 3,017 170 Journal of Vector Ecology June 2008

Table 2. Summary of data for fishes collected from abandoned swimming pools in New Orleans in January 2006.

# % of + % of collected # of Swimming Mean Catch Fishes Organisms Swimming (all species) Pools (n = 64) per + Pool* collected Pools Gambusia affinis 652 76.3% 16 25.0% 27.7 Poecillia latipinnia 81 9.5% 11 17.2% 2.7 Heterandria formosa 50 5.8% 9 14.1% 1.7 Cyprinodon varieagatus 42 4.9% 6 9.4% 1.4 Menidia beryllina 13 1.5% 1 1.6% 0.4 Lucania parva 8 0.9% 3 4.7% 0.3 Adinia xenica 4 0.5% 1 1.6% 0.1 Fundulus pulvereus 2 0.2% 2 3.1% 0.1 Cichlasoma cyanoguttatum 1 0.1% 1 1.6% <0.1 Fundulus chrysotus 1 0.1% 1 1.6% <0.1 Gobiosoma bosc 1 0.1% 1 1.6% <0.1 Total 855 * Mean Catch per + Pool refers to the average number of fish of each species collected in pools where at least one fish was collected. absence of the dominant winter mosquito Cs. inornata. Such to the walls and bottom of large deep containers may limit an increase in WNV vector abundance may increase WNV their efficacy in controlling mosquitoes. It is likely that the transmission to humans. Indeed members of our research odonate nymphs observed in our study have few encounters team recently demonstrated a significant increase in West with surface-breathing mosquitoes due to the depth and Nile neuroinvasive disease in 2006 in the hurricane-affected structurally simple walls of swimming pools. region (Caillouët et al. unpublished data). In many cases, the presence of competing predators or Fishes, consisting predominantly of G. affinis, were top predators has significant non-linear implications for the observed in about half of the swimming pools sampled. survival of prey (Wissinger and McGrady 1993, Finke and Mosquitofish Gambusia ( spp.) were used as the first Denno 2005). Because of the potential for predator-predator biological method of mosquito control in the early 20th antagonisms, trophic relationships should be evaluated prior century and have since been distributed worldwide (Bay to manipulating assemblages in biological control efforts 1967). The native range of G. affinis is the Mississippi River (Brodeur and Boivin 2006). Such antagonisms may result Basin (Ross 2001), and it is fortuitous that this species in the local extinction of mosquito predators when “top continues to perform its natural role in the very area where predator” species are also present (Marten et al. 2004), or it honed its mosquito hunting skills. Other fishes found in in “prey switching” in the presence of alternate prey sources swimming pools are commonly found in Lake Pontchartrain (Chesson 1989), or when competitors are present. Members and the drainage canals that flooded the Lakeview area of the Heteropteran families Belostomatidae and Nepidae (Duffy and Baltz 1998). No aquarium species were collected, are piscivorous as well as insectivorous. While these “top but a single non-native Rio Grande cichlid (O’Connell et al. predators” did not affect Odonata abundance in this study, 2002) was collected. Species richness of fishes was strongly they had a large negative effect onG. affinis abundance. This correlated to the proximity of the flooding source (the 17th resulted in an equally large positive indirect impact on the Street Canal) further implicating the levee breach as the presence of mosquitoes. However, this indirect effect was source of fishes in this study. less than half the magnitude of the direct effect of G. affinis Insect predators have received a great deal of interest on mosquito presence, suggesting that at the time of this as potential biological control agents due to their ability study G. affinis were effective in controlling mosquitoes to rapidly disperse into new habitats (Lacey and Orr 1994, despite the presence of strong top predators. Bay 1974). Of the predatory invertebrates we collected, Because abundance of native fishes had a large negative only odonate larvae appeared abundant enough to have a effect on mosquito abundance even with top predators and direct effect on immature mosquitoes. Nymphal odonates competitors present, manual distribution of G. affinis was are voracious predators of mosquito larvae in controlled suggested as an effective intervention at the time of this settings and small natural habitats (Fincke et al. 1997, Lacey study. Larval control of mosquitoes is an effective, cost- and Orr 1994, Stav et al. 2000, Yanoviak 2004). However, efficient method used to reduce mosquito populations and in our study, path analysis concluded that odonate nymphs the diseases they vector (Keiser et al. 2005). Based in part did not have a significant direct impact on mosquitoes on our findings, the New Orleans Mosquito and Termite in abandoned swimming pools. Sunahara et al. (2002) Control Board initiated a large-scale program of introducing proposed that the tendency for odonate nymphs to cling locally-collected G. affinis into abandoned pools throughout Vol. 33, no. 1 Journal of Vector Ecology 171

Figure 3. (A) Proposed model depicting expected direct effects (solid lines) and indirect effects (dashed lines) in a three-tiered food web with both specialist predators and other competitors present. The indirect effects (curved lines) of competitors and top predators on prey are expected to be positive due to direct suppression of specialist predators. Lines ending in arrows indicate a positive effect while those ending in circles indicate a negative effect. (B) Path diagram depicting observed direct (straight line) and indirect effects (curved line) of predators on the presence of mosquitoes in flooded swimming pools. Relationships that are statistically significant (P<0.05) are represented with solid lines. Widths of lines are proportional to the magnitude of effect with standardized coefficients next to each line. For this model, Chi-square = 0.152 with 1 degree of freedom resulting in a P-value of 0.696.

Orleans Parish during Spring 2006. Given the mitigating Carlson J., J. Keating, C.M. Mbogo, S. Kahindi, and J.C. effect that top predators and competitors had on mosquito Beier. 2004. Ecological limitations on aquatic mosquito presence via suppression of G. affinis, these pools will need predator colonization in the urban environment. J. to be periodically reevaluated to determine the need for Vector Ecol. 29: 331-339. additional control measures. Chesson, J. 1989. The effect of alternative prey on the functional response of Notonecta hoffmani. Ecology 70: Acknowledgments 1227-1235. Duffy, K.C. and D.M. Baltz. 1998. Comparison of fish We are very grateful for the field and lab support of assemblages associated with native and exotic Amanda Merkel, Frank Mannix, Gil Stav, Irene S. Caillouët, submerged macrophytes in the Lake Pontchartrain Jie Ting, Liz Hoy, and Dan Farrae. Thanks also to Greg estuary, USA. J. Exp. Marine Biol. Ecol. 223: 199-221. Thompson and Steve Sackett of the New Orleans Mosquito Durso, S.L. and M.J. Burguin. 1988. Mosquito abundance and Termite Control Board for their guidance and assistance and arboviral activity the Coachella Valley-1987. Proc. with this study. Finally, we would like to thank the residents Papers Annu. Conf. Calif. Mosq. Contr. Assoc. 56: 20- of the Lakeview and Filmore Neighborhoods for lending us 25. their swimming pools to study; we hope you will return soon Fincke, O.M., S.P. Yanoviak, and R.D. Hanschu. 1997. to clean out your pools and to be neighbors again. Stipend Predation by odonates depresses mosquito abundance support for this work was provided by the Louisiana Board in water-filled tree holes in Panama. Oecologia 112: of Regents Support Fund, Tulane University, and Centers 244-253. for Disease Control and Prevention Cooperative Agreement Finke, D.L. and R.F. Denno. 2005. Predator diversity and #T01/CCT622308. the functioning of ecosystems: the role of intraguild predation in dampening trophic cascades. Ecol. Lett. 8: REFERENCES CITED 1299–1306. Fox, M. and D.M. Wesson. 2005. Key to the Medically Bay, E.C. 1967. Mosquito control by fish: present-day Important Mosquitoes of Louisiana. Available from appraisal. Wld. Hlth. Org. Chron. 21: 415-423. http://www.cdc.gov/ncidod/dvbid/westnile/wnv_ Bay, E.C. 1974. Predator-prey relationships among aquatic hurricane.htm insects. Annu. Rev. Entomol. 19: 441-453. Goddard, L.B., A.E. Roth, W.K. Reisen, and T. Scott. 2002. Brodeur, J. and G. Boivin. 2006. Trophic and Guild Vector competence of California mosquitoes for West Interactions in Biological Control. Springer, New York Nile virus. Emerg. Infect. Dis. 9: 1385-1391. 249 p. Godsey, M.S., R. Nasci, H.M. Savage, S. Aspen, R. King, 172 Journal of Vector Ecology June 2008

A.M. Powers, K. Burkhalter, L. Colton, D. Charnetsky, stages of expansion for the introduced Rio Grande S. Lasater, V. Taylor, C.T. Palmisano. 2005. West Nile cichlid, Cichlasoma cyanoguttatum, in southeast Virus–infected mosquitoes, Louisiana, 2002. Emerg. Louisiana. Aquat. Invad. 13: 13-112. Infect. Dis. 11: 1399-1404. O’Malley, C.O. 1989. Guidelines for larval surveillance. Hammon W.M. and W.C. Reeves. 1943a. Laboratory Proceedings of the New Jersey Mosquito Control transmission of St. Louis encephalitis virus by three Association. genera of mosquitoes. J. Exp. Med. 78: 241–253. Plyer, A. and J. Bonaguro. 2007. Using U.S. Postal Service Hammon W.M. and W.C. Reeves. 1943b. Laboratory delivery statistics to track the repopulation of New transmission of western equine encephalomyelitis virus Orleans and the metropolitan area. Greater New by mosquitoes of the genera Culex and Culiseta. J. Exp. Orleans Community Data Center. Available from: Med. 78: 425–434. http://www.gnocdc.org/reports/GNOCDC_research_ Keiser, J., B.H. Singerb, and J. Utzinger. 2005. Reducing the note_May07.pdf burden of malaria in different eco-epidemiological Reichart, T., E. Check, and E. Marris. 2005. After the flood. settings with environmental management: a systematic Nature 437: 174-176. review. Lancet Infect. Dis. 5: 695-708. Ross, S. 2001. The Inland Fishes of Mississippi. University Lacey, L.A. and B.K. Orr. 1994. The role of biological control Press of Mississippi, 376 p. of mosquitoes in integrated vector control. Am. J. Trop. Shultz, J.M., J. Russell, and Z. Espinel 2005. Epidemiology of Med. Hyg. 50: 97-115. tropical cyclones: the dynamics of disaster, disease, and Marten, G.G., E.S. Bordes, and M. Nguyen. 2004. development. Epidemiol. Rev. 27: 21-35. Use of cyclopoid copepods for mosquito control. Stav, G., L. Blaustein, and Y. Margalit. 2000. Influence of Hydrobiologia 292: 491-496. nymphal Anax imperator (Odonata: Aeshnidae) on Manickchand-Heileman, S., J. Mendoza-Hill, and A.L. oviposition by the mosquito Culiseta longiareolata Kong. 2004. A trophic model for exploring possible (Diptera: Culicidae) and community structure in ecosystem impacts of fishing in the Gulf of Paria, temporary pools. J. Vector Ecol. 25: 190-202. between Venezuela and Trinidad. Ecol. Model. 172: Sunahara, T., K. Ishizaka, and M. Mogi. 2002. Habitat size: 307–322. a factor determining the opportunity for encounters Merritt, R.W. and K.W. Cummings. 1996. An Introduction between mosquito larvae and aquatic predators. J. to the Aquatic Insects of North America. Kendall Hunt Vector Ecol. 27: 8-20. Publishing Co. Dubuque, IA. Townsend, J. 2005. West Nile Virus in Maricopa County Michaels, S.R., G.A. Balsamo, M. Kukreja, C.H. Anderson, 2004. Presentation in Sixth National Conference on S. Straif-Bourgeois, G. Talati, and R.C. Ratard. 2005. West Nile Virus in the United States. CDC. Available Surveillance for West Nile virus cases in Louisiana online: http://www.cdc.gov/ncidod/dvbid/westnile/ 2001-2004. J. LA State Med. Soc. 157: 269-272. conf/February_2005.htm Miura, T. and R.M. Takahashi. 1988. A laboratory study of Turner, A.M. and J.C. Trexler. 1997. Sampling aquatic predation by damselfly nymphs, Enallagma civile, upon invertebrates from marshes: evaluating the options. J. mosquito larvae, Culex tarsalis. J. Am. Mosq. Contr. N. Am. Benthol. Soc. 16: 694-709. Assoc. 4: 129-131. Wissinger, S. and J. McGrady. 1993. Intraguild predation Nasci R.S. and C.G. Moore. 1998. Vector-borne disease and competition between larval dragonflies: direct and surveillance and natural disasters. Emerg. Infect. Dis. indirect effects on shared prey. Ecology 74: 207-218. 4: 333-334. Yanoviak, S.P. 2004. Predation, resource availability, and O’Connell, M., R.C. Cashner, and G. Fuentes. 2002. community structure in Neotropical water-filled tree Application of the diffusion model to describe a recent holes. Oecologia 126: 125-133. invasion: Observations and insights concerning early